One of the single most important opportunities to influence an individual's health is through sustained modification of diet. Although certain dietary patterns have clear health benefits or detriments with respect to chronic disease risk, the biochemical and metabolic underpinnings linking such dietary patterns with health effects are not clear. This proposal describes a step-wise strategy whereby the mechanistic effects of individual foods and dietary components on health can be delineated as follows: Step 1) identify molecular signatures of diet/food composites from typical Mediterranean-style (MED) and Westernized diets (WD); Step 2) identify biomarkers of MED and WD exposures in human biospecimens (e.g., plasma, urine, stool); Step 3) connect these signatures and biomarkers with health indicators, including gut microbiome, in human subjects; and Step 4) test these associations in a prospective clinical trial. This proposal leverages already available specimens (blood, urine, stool) from two randomized, controlled feeding trials of MED and WD. Measures of lipid profile, blood pressure, insulin sensitivity, inflammation, and other health indicators are available and will be associated with the comprehensive molecular metabolic profiles generated in this proposal. The central hypothesis is that different dietary patterns (MED vs. WD) will generate distinct food-specific molecular signatures that can be measured in human biospecimens and linked to metabolic health indicators. Guided by strong preliminary data, our hypotheses will be tested in the following specific aims (SA) by using state-of-art foodomics, plasma/urine metabolomics, targeted analysis of known nutrients, and fecal metabolomic/metagenomic approaches to identify and validate molecular signatures in foods and biomarkers in biospecimens (SA 1 and 2); by linking these signatures and biomarkers with health indicators and the gut microbiota (SA 3); and by testing the specific foods or food groups identified in Aims 1-3 that are beneficial to metabolic health in a randomized, prospective, controlled feeding trial (SA 4). The proposed research is significant because, for the first time, it will use a step-wise approach to directly connect a large panel of individual foods and whole diets to metabolic and intestinal health indicators through molecular and metabolic signatures. Results can be used almost immediately to inform future studies. The approach is innovative because it represents a substantial departure from the status quo by examining the nutritional metabolome as part of the broader ?exposome? and identifying the overlap between individual foods, composite diets, and biospecimens. Findings from this research will elucidate relationships between dietary exposures and metabolic health by identifying and connecting unique food signatures and biological metabolites. This knowledge will underpin the development of effective and more precise dietary interventions for metabolic disease prevention and recommendations to promote health.